Lifepo4 cells catch fire!

[Grumpy-b]

Fused and internally protected. Still fully functioning. The connection clip appears to have come off when pack collapsed. So I guess the power slowly built up heat until the thermal runaway took over. Thankfully the Valence chemistry and construction makes them relatively safe in such a situation, and very unlikly to burst into flames. For that I am eternally grateful.

Grumpy-b

[GregsGarage]

Their plastic spacer blocks don't keep them close together but this ruins the density and makes for a big pack.

That's what I thought until I did some calculations of pack size. What I found is that the Headway cells would make for a smaller pack not larger as you would expect.

[GregsGarage]

Interestingly even with a BMS, if one cell had internally shorted, i doubt if anything would have been simply detected.

I am going to disagree. You measured around 10 volts before you started. This doesn't tell you if it is safe to charge. To find this out you have to measure all cell voltages. In this case, you only need to take 4 measurements, since parallel connected cells all have the same voltage. I suspect you would have found something like 3 at 3.3 volts and 1 close to 0. In this condition it is unsafe to charge. If you had all 4 measurements around 2.5 volts then this would be safe for charging, but you want to periodically retake these measurments during the charge. This is where a bms can help. It automates this process and should have control over the charger so it can shut it off. Your pictures do show some sort of bms board built into the cells, but I am guessing that you didn't have it connected to your charger. I have not used the Valence product and don't know if you can easily interface with the cells built-in bms circuitry or have to purchase a interface from them. Some sort of battery managment strategy would have prevented this.

Another view on why the fire wasn't worse. Your 4 amp charger into a 130 ah pack was only a third of the charged rate relative to pack capacity as my 10 amp charger into a 100 ah pack. It is possible that had you been using a 13 amp charger it may have been much worse.

[Grumpy-b]

You cant check each cell its a 12v equivalent Battery case thats sealed and unservicable. With internal monitoring circuit. So its totally impossible to know what each cell is doing. Its quite possible that if one has the Valence diagnostic software and connector that you could interrogate the pack.
I dont have that, and neither will many people in the future, so others getting hold of such items could readily do what I did. Not justifying my self, I guess I should have just thought I have i have $8to 10K worth of batteries here they may be of some use but I should just get rid of them to be sure. Hmmm like thats going to happen.
These are sold as being a 12v replacement product. Hence the specific form factor of the casings. It would almost certainly need to have been told it was a single battery not part of a string.Perhaps a default if not daisychained together that makes that so could be an advantage.

Seeing how they are made all of the cells get the same voltage there is no internal shunting etc to make any part of the battery receive any different voltage.

Grumpy-b

[Beemer]

Their plastic spacer blocks don't keep them close together but this ruins the density and makes for a big pack.

That's what I thought until I did some calculations of pack size. What I found is that the Headway cells would make for a smaller pack not larger as you would expect.

Check out Damiens website: http://www.evbmw.com/
His headways fill his boot and gives him a 30 mile range.

He's looking at A123, 20AH's now. Said they scare him silly. One melted his dummy load at 400 amps.

[GregsGarage]

His headways fill his boot and gives him a 30 mile range.

The closest equivalent prismatic pack would be 50x 60ah cells. That would also fill his boot.

[Beemer]

His headways fill his boot and gives him a 30 mile range.

The closest equivalent prismatic pack would be 50x 60ah cells. That would also fill his boot.

http://en.calb.cn/Product/?id-116.html

50x60AH Calb's are:-
44mm*243mm*115mm*50off = 0.0615m^3 = 1.65cu foot.
It's only 9.6KWH max. CALB's are renowned for good quality control. you should expect 10KWH and all the cells to be all the same capacity etc.

On second thoughts I should of bought these over A123's.
Nope, sorry matey, cannot agree. Maybe you have the old sizes?

[GregsGarage]

Using your sizes a single 60ah cell is 1280cm^3.
The 16ah headways measure 40mm x 170mm. So 4 of them give you 64ah. Arrange them in a square gives you 8 x 8 x 17cm = 1088cm^3. Less volume, more capacity.

[Beemer]

Using your sizes a single 60ah cell is 1280cm^3.
The 16ah headways measure 40mm x 170mm. So 4 of them give you 64ah. Arrange them in a square gives you 8 x 8 x 17cm = 1088cm^3. Less volume, more capacity.

Headway 40152 4.5 * 4.5 * 17 * 4 = 1377
CALB CA 4.4 * 11.5 * 24.3 = 1230
The latter wins by 12%.

I've placed 4.5cm across cells centres because of the gap between the cells on the standard "square" connectors. Headway energy density is CALB/Sinopoly standard at 105Wh/Kg. make sure you are buying the newer Headways, the older ones used to fail on their own with a suspected bad batch of electrolyte (forum speak). Smaller cells heat up faster than large cells. That means disasters happen with less warning.

However, there are Headway spacers which place the cells into a triangle so a little bit is gained back.
On the other hand... The CALB SE series being older are larger, the capacity is the same 67AH as the smaller CA series which seems to lose a little less Voltage under load.

What would do it for me in a car would be quality against price. I bought A123's for power density against price (and the challenge).

Go for it matey. Damien is enjoying his

[GregsGarage]

I got the volume for the calb cells wrong, it should have been 1230cc as you say. However your spacings for using the Headway spacer blocks is 2mm, not 5mm. I've got some for testing. Also Damien is using the 40160s 16ah cells. The 40152 cells are shorter and have less capacity. So the correct figure for 40160 headway cells with their spacer blocks is 4.2 x 4.2 x 17 x 4 = 1200cc. Still slightly less volume than the equivalent 60ah CALB with slighty more manufacturer's specified capacity.

So while you would think the prismatic cells would be better than cylindrical cells regards to volume, actually the cylindrical cells are better, although there is not much in it. This can be easily explained because the prismatics have a thick plastic case and also more dead space for internal cell connections. The headways have a thin metal case and need much less space for the internal connections to the external cell terminals.

Another thing that has impressed me with Damian's setup is how well his cells have held up to what he has dished out to them. Many prismatics would have died long ago with the high C rates they have had to deal with.